*** Replication Code for "Who Shapes the Law? Gender and Racial Bias in Judicial Citations"
*** APSR, Szmer, Moyer, Haire, and Christensen
*** Estimated using Stata 17.0
*** Note: The coefplot command used to generate some of the figures is an .ado file. Use the following command install the file
**** findit coefplot

* We recommend using the cd, or change directory command here


use "SMHC_Who_Shapes_the_Law.dta", clear
****** In 240 case comparison of DV with a dV starting 3 months after the decision, initial coder appeared to include a leading 1 (11); in the second, a leading 2 (23)
*replace citation_5yr_diff_circ_2022=1 if casenum ==26254
*replace citation_5yr_diff_circ_2022=3 if casenum ==25932

********************************************* Figures in the text *********************************************
**** Figure 1: Race & Gender's Effect on Out-of-Court Citations
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct i.year if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)

estimates store model1_no_interact
 coefplot (model1_no_interact, mcolor(black) ciopts(lcol(gray))), grid(none) keep(OAminority femaleOA) graphregion(fcolor(white) lcolor(black) lwidth(medium)) plotregion(fcolor(white)) xline(0, lcolor(black))  xscale(range(-0.6 0.40)) xlabel(-0.6(0.25)0.40) title("") coeflabels(femaleOA= "Female Judge" OAminority="Non-white Judge") title("Figure 1: Effect of Race and Gender on Out-of-Circuit Citations", size(11pt) color(black))
graph save apsr_figure1, replace
graph export apsr_figure1.tiff, as(tif) replace
**** Figure 2: How Race and Gender Condition the Effect of Opinion Quality on Out of Court Citations
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority authorities_deep  femaleOA_deep_cites OAminority_deep_cites splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct i.year if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1 & OAdist_ct~=1, vce(cluster circuit)

estimates store model2_interact

nbreg citation_5yr_diff_circ_2022  femaleOA OAminority authorities_deep OAminority_deep_cites splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1 & OAdist_ct~=1, vce(cluster circuit)

estimates store model3_interact

 coefplot (model2_interact, mcolor(black) ciopts(lcol(gray))), grid(none) keep(OAminority femaleOA authorities_deep femaleOA_deep_cites OAminority_deep_cites) graphregion(fcolor(white) lcolor(black) lwidth(medium)) plotregion(fcolor(white)) xline(0, lcolor(black))  xscale(range(-0.6 0.40)) xlabel(-0.6(0.25)0.40) title("")
graph save model_2_both_interacts, replace 

 coefplot (model3_interact, mcolor(black) ciopts(lcol(gray))), grid(none) keep(OAminority femaleOA authorities_deep OAminority_deep_cites) graphregion(fcolor(white) lcolor(black) lwidth(medium)) plotregion(fcolor(white)) xline(0, lcolor(black))  xscale(range(-0.6 0.40)) xlabel(-0.6(0.25)0.40) title("")
graph save model_3_best, replace
graph combine model_2_both_interacts.gph model_3_best.gph, col(1)  graphregion(fcolor(white) lcolor(black) lwidth(medium)) plotregion(fcolor(white)) title("Figure 2: How Race and Gender Condition the Effect of Opinion Quality on" "Out-of-Circuit Citations", size(11pt) color(black)) 
graph export apsr_figure2.tiff, as(tif) replace

**** Figure 3: Expected Counts of Subsequent External Citations to Opinions Written by Non-white and White Judges
nbreg citation_5yr_diff_circ_2022  i.femaleOA c.authorities_deep##i.OAminority splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct i.year if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1 & OAdist_ct~=1, vce(cluster circuit)
margins OAminority, at(authorities_deep=(0(1)52))
marginsplot, ytitle(Expected Number of Subsequent Cites)  noci scheme(s1mono) graphregion(fcolor(none) ifcolor(none)) plotregion(fcolor(none))  recast(line) legend(order(1 "White Opinion Author" 2 "Non-white Opinion Author")) title("Figure 3: Expected Counts of Out-of-Circuit Citations to Opinions" "Written by Non-White and White Judges", size(11pt) color(black)) 
graph export apsr_figure3.tiff, as(tif) replace

*** Percent Change in Expected Count for Women v. Men (mentioned in discussion)--Using Model 3, best model
qui nbreg citation_5yr_diff_circ_2022 i.femaleOA i.OAminority##c.authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1 & OAdist_ct~=1, vce(cluster circuit)

nbreg citation_5yr_diff_circ_2022 i.femaleOA OAminority authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct i.year if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
* women on average receive 0.251 fewer cites than men. Keeping in mind that the average number of cites is 1.544 over the subsequent five years. 
margins, at(femaleOA=(0 1)) post
lincom _b[2._at] - _b[1._at], 
display ((_b[2._at] - _b[1._at])/_b[1._at])*100 

**** Intersectionality Models
*gen OA_minority_women=0
 *replace OA_minority_women=1 if femaleOA==1 & OAminority==1
 *lab var OA_minority_women "Non-white Female Judge"
 *gen OA_minority_man=0
 *replace OA_minority_man=1 if femaleOA==0 & OAminority==1
*lab var OA_minority_man "Non-white Male Judge"
 *gen OA_white_woman=0
  *replace OA_white_woman=1 if femaleOA==1 & OAminority==0
  *lab var OA_white_woman  "White Female Judge"
* gen OAminorwomen_deepcites=OA_minority_women*authorities_deep
* lab var OAminorwomen_deepcites "Non-white Female Judge X Deep Cites"
* gen OAminormen_deepcites=OA_minority_man*authorities_deep
* lab var OAminormen_deepcites "Non-white Male Judge X Deep Cites" 
* gen OAwhitewomen_deepcites=OA_white_woman*authorities_deep
* lab var OAwhitewomen_deepcites "White Female Judge X Deep Cites"

*** Figure 4: Intersectionality Analysis of the Effects of Race and Gender
  
  nbreg citation_5yr_diff_circ_2022 OA_minority_women OA_minority_man OA_white_woman  authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)


estimates store intersect_no_interact
 coefplot (intersect_no_interact, mcolor(black) ciopts(lcol(gray))), grid(none) keep(OA_minority_women OA_minority_man OA_white_woman) graphregion(fcolor(white) lcolor(black) lwidth(medium)) plotregion(fcolor(white)) xline(0, lcolor(black))  xscale(range(-0.6 0.40)) xlabel(-0.6(0.25)0.40) title("") coeflabels(OA_minority_women= "Non-white Female Judge" OA_minority_man="Non-white Male Judge" OA_white_woman="White Female Judge") title("Figure 4: Effects of Judge Race-Gender on Out-of-Circuit" "Citations", size(11pt) color(black))
graph save apsr_figure4.gph, replace
graph export apsr_figure4.tiff, as(tif) replace

*** pct change for nonwhite women; 28% fewer subsequent citations for an individual case, when compared to a white male; the difference is 0.446, or almost half a case...
*** that, of course adds up (replicate using the same formula for the number we put in the discussion for women in genera;)
  nbreg citation_5yr_diff_circ_2022 OA_minority_women OA_minority_man OA_white_woman  authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)

  
margins, at(OA_minority_women=(0 1)) post
lincom _b[2._at] - _b[1._at], 
display ((_b[2._at] - _b[1._at])/_b[1._at])*100 
*** 

*** using that to generate the projected differential in cites over 10 year period.... So, per year, each active judge from 2011-2016 FY wrote on, average. 53.5 signed opinions. (https://www.uscourts.gov/file/22024/download)
*** for each year, they would write 53.5 signed opinions. Over a 5 year period, white males would have 53.5 X 0.446=23.86 more subsequent cites for the opinions written in that one year. Over 10 years, that would result in 238.6 (or 239) more citations.

*** Figure 5: Intersectionality and Conditional Effect of Opinion Quality
 **** Interacting the dummies with the authorities deep
  nbreg citation_5yr_diff_circ_2022 i.OA_minority_women##c.authorities_deep i.OA_minority_man##c.authorities_deep i.OA_white_woman##c.authorities_deep  authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)
  

  nbreg citation_5yr_diff_circ_2022 OA_white_woman OA_minority_women OA_minority_man authorities_deep OAminorwomen_deepcites OA_minority_man OAminormen_deepcites OAwhitewomen_deepcites splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)
estimates store intersect_model2

  *** interacting the minority dummies only with authorities deep
    nbreg citation_5yr_diff_circ_2022 i.OA_minority_women##c.authorities_deep i.OA_minority_man##c.authorities_deep i.OA_white_woman   authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)
  

  nbreg citation_5yr_diff_circ_2022 OA_white_woman OA_minority_women OA_minority_man authorities_deep OAminorwomen_deepcites OA_minority_man OAminormen_deepcites splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)
 estimates store intersect_model3

 coefplot (intersect_model2, mcolor(black) ciopts(lcol(gray))), grid(none) keep(OA_minority_women OA_minority_man OA_white_woman authorities_deep OAminorwomen_deepcites OAminormen_deepcites OAwhitewomen_deepcites) graphregion(fcolor(white) lcolor(black) lwidth(medium)) plotregion(fcolor(white)) xline(0, lcolor(black))  xscale(range(-0.6 0.40)) xlabel(-0.6(0.25)0.40) title("")
graph save APSR_intersect_model2.gph, replace 

 coefplot (intersect_model3, mcolor(black) ciopts(lcol(gray))), grid(none) keep(OA_white_woman OA_minority_women OA_minority_man  authorities_deep OAminorwomen_deepcites OAminormen_deepcites) graphregion(fcolor(white) lcolor(black) lwidth(medium)) plotregion(fcolor(white)) xline(0, lcolor(black))  xscale(range(-0.6 0.40)) xlabel(-0.6(0.25)0.40) title("")
graph save APSR_intersect_model3.gph, replace
graph combine APSR_intersect_model2.gph APSR_intersect_model3.gph, col(1)  graphregion(fcolor(white) lcolor(black) lwidth(medium)) plotregion(fcolor(white)) title("Figure 5: How Judge Race-Gender Conditions the" "Effect of Opinion Quality on Out-of-Circuit Citations", size(11pt) color(black)) 
graph save apsr_figure5.gph, replace
graph export apsr_figure5.tiff, as(tif) replace
 
***** Figure 6: Marginal Effects Graph, intersectionality using the model with both interactions (model 2 for intersectionality)
nbreg citation_5yr_diff_circ_2022  i.OA_minority_women##c.authorities_deep i.OA_minority_man##c.authorities_deep    i.OA_white_woman##c.authorities_deep  splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)

margins, at(authorities_deep=(0(1)52)) over(OA_minority_women OA_minority_man OA_white_woman)


marginsplot, noci plot1opts(msymbol(circle) mcolor(gray) mfcolor(white)) plot2opts(msymbol(diamond) mcolor(gray)) plot3opts(msymbol(point) mcolor(black) mfcolor(black)) plot4opts(msymbol(square) mcolor(gray) mfcolor(white)) graphregion(fcolor(white)) ytitle("Expected Number of Subsequent Cites", size(medium) margin(medium)) ylabel(, labsize(medium)) legend(order(1 "White Male" 2 "White Woman" 3 "Non-white Male" 4 "Non-white Woman") size(medsmall) color(black)) title("Figure 6: Expected Counts of" "Out-of-Circuit Citations by Judge Race-Gender", color(black))  scheme(s1mono) graphregion(fcolor(none) ifcolor(none)) plotregion(fcolor(none))  

graph export apsr_figure6.tiff, as(tif) replace
 

********************************************* Supplemental Materials (Online Appendix) *********************************************

**** OA 1. Discussion of Variable Coding & Descriptive Statistics
**** Table OA1.1: Descriptive Statistics
nbreg citation_5yr_diff_circ_2022 femaleOA  OA_white_woman authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct i.year if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1 & OAdist_ct~=1, vce(cluster circuit)
estpost sum citation_5yr_diff_circ_2022 femaleOA OAminority OA_white_woman OA_minority_women OA_minority_man authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct if est_sample==1


esttab  using OA1_table_descriptives_APSR.rtf, cell((mean(label(Mean)) sd(label(Std. Dev.)) min(label(Minimum)) max(label(Maximum)))) modelwidth(10) varwidth(29) label nonumbers replace

**** OA2. Full Results of the Models in Figures 1, 2, 4, and 5
**** Table OA2.1: Negative Binomial Regression of the Number of Subsequent Citations Outside of the Circuit 
**** NOTE: To facilitate the esttab command, we use the actual dummy variable coding for year fixed effects and we manually coded the multiplicative terms for Female Judge X Deep Cites and Non-white Judge X Deep Cites

****** Model 1: No Interaction
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)
estimates store model1_no_interact
****** Model 2: Interacting Non-white Judge and Gender with Deep Cites
 nbreg citation_5yr_diff_circ_2022 femaleOA OAminority femaleOA_deep_cites OAminority_deep_cites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
estimates store model2_interact
****** Model 3: Interacting Non-white Judge with Deep Cites 
 nbreg citation_5yr_diff_circ_2022 femaleOA OAminority OAminority_deep_cites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
estimates store model3_interact

esttab model1_no_interact  model2_interact model3_interact using OA2Table1_APSR.rtf, pr2 b(%20.3f) se(%12.3f) pr2(%9.3f)  varwidth(27) modelwidth(8) label coeflabels(_cons "Constant") star(* 0.05 ** 0.01 *** 0.001) order(femaleOA femaleOA_deep_cites OAminority OAminority_deep_cites authorities_deep) replace drop(yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016) mtitles("Model 1" "Model 2" "Model 3") nonumbers

******* Table OA2.2: Negative Binomial Regression of the Number of Subsequent Citations Outside of the Circuit: Race-Gender Groups 
nbreg citation_5yr_diff_circ_2022 OA_white_woman OA_minority_women OA_minority_man authorities_deep  splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)
estimate store no_interact_intersect

nbreg citation_5yr_diff_circ_2022  OA_minority_women OA_minority_man OA_white_woman OAminorwomen_deepcites OAminormen_deepcites OAwhitewomen_deepcites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)
estimate store interact_intersect


nbreg citation_5yr_diff_circ_2022  OA_minority_women OA_minority_man OA_white_woman OAminorwomen_deepcites OAminormen_deepcites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)


estimate store interact_intersect_nonwhite

esttab no_interact_intersect  interact_intersect interact_intersect_nonwhite using OA2Table2_APSR.rtf, pr2 b(%20.3f) se(%12.3f) pr2(%9.3f)  varwidth(27) modelwidth(8) label coeflabels(_cons "Constant") star(* 0.05 ** 0.01 *** 0.001) order(OA_minority_women OA_minority_man OA_white_woman OAminorwomen_deepcites OAminormen_deepcites OAwhitewomen_deepcites authorities_deep) replace drop(yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016) mtitles("Model 4" "Model 5" "Model 6") nonumbers

**** OA3. Alternative Specifications Including Asian-American Judges
***** Table OA3.1: NBR Model of Citations, Asian-American Judges Coded as Non-White
********** Model 1: No Interaction
nbreg citation_5yr_diff_circ_2022 femaleOA j1_minority authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & OAdist_ct~=1 , vce(cluster circuit)
estimates store model1_OA_asian_min
********** Model 2: Interacting Black/Latinx/Asian Judge and Gender with Deep Cites
 nbreg citation_5yr_diff_circ_2022 femaleOA j1_minority femaleOA_deep_cites j1_minority_deepcites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & OAdist_ct~=1, vce(cluster circuit)
estimates store model2_OA_asian_min
********** Model 3: Interacting Black/Latinx/Asian Judge with Deep Cites
 nbreg citation_5yr_diff_circ_2022 femaleOA j1_minority j1_minority_deepcites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & OAdist_ct~=1, vce(cluster circuit)
estimates store model3_OA_asian_min

esttab model1_OA_asian_min  model2_OA_asian_min model3_OA_asian_min using OA3_APSR_asian_min.rtf, pr2 b(%20.3f) se(%12.3f) pr2(%9.3f)  varwidth(33) modelwidth(8) label coeflabels(_cons "Constant" ) star(* 0.05 ** 0.01 *** 0.001) order(femaleOA femaleOA_deep_cites j1_minority j1_minority_deepcites authorities_deep) replace drop(yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016) mtitles("Model 1" "Model 2" "Model 3") nonumbers

**** OA4: Alternative Specifications with Control for Federalist Society Affiliation    
lab var OA_fed_soc_national_covention "FedSoc National Convention"

******Table OA4.1: NBR Model of Citations, Control for Federalist Society (Hollis-Brusky measure)
********** Model 1: No Interaction
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority authorities_deep OA_fed_soc_national_covention splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)
estimates store model1_fedsoc_hb
********** Model 2: Interacting Non-white Judge and Gender with Deep Cites
 nbreg citation_5yr_diff_circ_2022 femaleOA OAminority femaleOA_deep_cites OAminority_deep_cites authorities_deep OA_fed_soc_national_covention splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
estimates store model2_fedsoc_hb
********** Model 3: Interacting Non-white Judge with Deep Cites 
 nbreg citation_5yr_diff_circ_2022 femaleOA OAminority OAminority_deep_cites authorities_deep OA_fed_soc_national_covention splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
estimates store model3_fedsoc_hb

esttab model1_fedsoc_hb  model2_fedsoc_hb model3_fedsoc_hb using OA4_fedsoc_HB_APSR.rtf, pr2 b(%20.3f) se(%12.3f) pr2(%9.3f)  varwidth(27) modelwidth(8) label coeflabels(_cons "Constant") star(* 0.05 ** 0.01 *** 0.001) order(femaleOA femaleOA_deep_cites OAminority OAminority_deep_cites authorities_deep) replace drop(yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016) mtitles("Model 1" "Model 2" "Model 3") nonumbers



**** OA5. Exploring Trends over Time

*** Figure OA5.1: Coefficient Plots of the Marginal Effects of Race/Gender on Subsequent Cites, by Year
*
nbreg citation_5yr_diff_circ_2022 i.femaleOA##c.yr_count i.OAminority##c.yr_count authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct  if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
margins, dydx(femaleOA) by(yr_count) post
estimates store yrcount_fem
coefplot (yrcount_fem, mcolor(black) ciopts(lcol(gray))), grid(none) graphregion(fcolor(white) lcolor(black) lwidth(medium)) plotregion(fcolor(white)) nokey xtitle("Effect of Female Judge, by Year", size(medium)) ytitle("") ylabel(1 "2009" 2 "2010" 3 "2011" 4 "2012" 5 "2013" 6 "2014" 7 "2015" 8 "2016", angle() labsize(medium)) yscale(lcolor(black)) xline(0, lcolor(black))  xscale(range(-1.1 1.1)) xlabel(-1.1(0.20)1.1) saving(APSR_OA5_yrcount_female.gph, replace)


nbreg citation_5yr_diff_circ_2022 i.femaleOA##c.yr_count i.OAminority##c.yr_count authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct  if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
margins, dydx(OAminority) by(yr_count) post
estimates store yrcount_minority
coefplot (yrcount_minority, mcolor(black) ciopts(lcol(gray))), grid(none) graphregion(fcolor(white) lcolor(black) lwidth(medium)) plotregion(fcolor(white)) nokey xtitle("Effect of Non-White Judge, by Year") ytitle("") ylabel(1 "2009" 2 "2010" 3 "2011" 4 "2012" 5 "2013" 6 "2014" 7 "2015" 8 "2016", angle() labsize(medium)) yscale(lcolor(black)) xline(0, lcolor(black))  xscale(range(-1.1 1.1)) xlabel(-1.1(0.20)1.1) saving(APSR_OA5_yrcount_nonwhite.gph, replace)

nbreg citation_5yr_diff_circ_2022 i.OA_white_woman##c.yr_count i.OA_minority_women##c.yr_count i.OA_minority_man##c.yr_count authorities_deep  splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
margins, dydx(OA_white_woman) by(yr_count) post
estimates store yrcount_whitefem
coefplot (yrcount_whitefem, mcolor(black) ciopts(lcol(gray))), grid(none) graphregion(fcolor(white) lcolor(black) lwidth(medium)) plotregion(fcolor(white)) nokey xtitle("Effect of White Female Judge, by Year") ytitle("") ylabel(1 "2009" 2 "2010" 3 "2011" 4 "2012" 5 "2013" 6 "2014" 7 "2015" 8 "2016", angle() labsize(medium)) yscale(lcolor(black)) xline(0, lcolor(black))  xscale(range(-1.1 1.1)) xlabel(-1.1(0.20)1.1) saving(APSR_OA5_yrcount_whitefem.gph, replace)

nbreg citation_5yr_diff_circ_2022 i.OA_white_woman##c.yr_count i.OA_minority_women##c.yr_count i.OA_minority_man##c.yr_count authorities_deep  splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct  if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
margins, dydx(OA_minority_women) by(yr_count) post
estimates store yrcount_minorfem
coefplot (yrcount_minorfem, mcolor(black) ciopts(lcol(gray))), grid(none) graphregion(fcolor(white) lcolor(black) lwidth(medium)) plotregion(fcolor(white)) nokey xtitle("Effect of Non-white Female Judge, by Year") ytitle("") ylabel(1 "2009" 2 "2010" 3 "2011" 4 "2012" 5 "2013" 6 "2014" 7 "2015" 8 "2016", angle() labsize(medium)) yscale(lcolor(black)) xline(0, lcolor(black))  xscale(range(-1.1 1.1)) xlabel(-1.1(0.20)1.1) saving(APSR_OA5_yrcount_nonwhitefem.gph, replace)

nbreg citation_5yr_diff_circ_2022 i.OA_white_woman##c.yr_count i.OA_minority_women##c.yr_count i.OA_minority_man##c.yr_count authorities_deep  splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct  if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
margins, dydx(OA_minority_man) by(yr_count) post
estimates store yrcount_minormale
coefplot (yrcount_minormale, mcolor(black) ciopts(lcol(gray))), grid(none) graphregion(fcolor(white) lcolor(black) lwidth(medium)) plotregion(fcolor(white)) nokey xtitle("Effect of Non-white Male Judge, by Year") ytitle("") ylabel(1 "2009" 2 "2010" 3 "2011" 4 "2012" 5 "2013" 6 "2014" 7 "2015" 8 "2016", angle() labsize(medium)) yscale(lcolor(black)) xline(0, lcolor(black))  xscale(range(-1.1 1.1)) xlabel(-1.1(0.20)1.1) saving(APSR_OA5_yrcount_nonwhiteman.gph, replace)


graph combine APSR_OA5_yrcount_female.gph APSR_OA5_yrcount_nonwhite.gph APSR_OA5_yrcount_whitefem.gph APSR_OA5_yrcount_nonwhitefem.gph APSR_OA5_yrcount_nonwhiteman.gph, graphregion(color(white)) row(3)  saving(APSR_OA5_figure1, replace) 

graph export "APSR_OA5_Figure1.tif", replace

********* Dataverse Appendix ****************
**** DA1: Illustrating the Average Marginal Effects of Non-white Judge by Observed Values of Deep Cites
**** Figure DA1.1: Illustrating the Average Marginal Effects of Non-white Judge by Observed Values of Deep Cites
qui nbreg citation_5yr_diff_circ_2022  i.femaleOA c.authorities_deep##i.OAminority splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct i.year if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1 & OAdist_ct~=1, vce(cluster circuit)

margins, dydx(OAminority) at(authorities_deep=(0(1)52)) 
marginsplot, ytitle("Marginal Effect of Non-white Opinion Author") title("")  scheme(s1mono) graphregion(fcolor(none) ifcolor(none)) plotregion(fcolor(none)) yline(0) recastci(rarea)  recast(line)
graph save apsr_figure_DA1_1.gph, replace
graph export apsr_figure_DA1_1.tiff, as(tif) replace

**** Figure DA1.2 Illustrating the Average Marginal Effects of Non-white Women Judge by Observed Values of Deep Cites
nbreg citation_5yr_diff_circ_2022  i.OA_minority_women##c.authorities_deep i.OA_minority_man##c.authorities_deep    i.OA_white_woman  splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)

margins, dydx(OA_minority_women) at(authorities_deep=(0(1)52))
marginsplot, ytitle("Marginal Effect of Non-white Women Opinion Authors") title("")  scheme(s1mono) graphregion(fcolor(none) ifcolor(none)) plotregion(fcolor(none)) yline(0) recastci(rarea)  recast(line)

graph save apsr_figure_DA1_2.gph, replace
graph export apsr_figure_DA1_2.tiff, as(tif) replace
**** Figure DA1.3 Illustrating the Average Marginal Effects of Non-white Male Judge by Observed Values of Deep Cites

nbreg citation_5yr_diff_circ_2022  i.OA_minority_women##c.authorities_deep i.OA_minority_man##c.authorities_deep    i.OA_white_woman  splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)
margins, dydx(OA_minority_man) at(authorities_deep=(0(1)52))
marginsplot, ytitle("Marginal Effect of Non-white Male Opinion Authors") title("")  scheme(s1mono) graphregion(fcolor(none) ifcolor(none)) plotregion(fcolor(none)) yline(0) recastci(rarea)  recast(line)

graph save apsr_figure_DA1_3.gph, replace
graph export apsr_figure_DA1_3.tiff, as(tif) replace


**** DA2. Additional Alternative Specifications Including Asian-American Judges

****** Table DA2.1: NBR Model of Citations, Asian-American Judges Included in Reference 
********** Model 1: No Interaction
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & OAdist_ct~=1 , vce(cluster circuit)
estimates store model1_OA_asian_ref
********** Model 2: Interacting Black or Latinx Judge and Gender with Deep Cites
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority femaleOA_deep_cites OAminority_deep_cites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & OAdist_ct~=1, vce(cluster circuit)
estimates store model2_OA_asian_ref
********** Model 3: Interacting Black or Latinx Judge with Deep Cites 
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority OAminority_deep_cites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & OAdist_ct~=1, vce(cluster circuit)
estimates store model3_OA_asian_ref

esttab model1_OA_asian_ref  model2_OA_asian_ref model3_OA_asian_ref using DA2_1_APSR_asian_ref.rtf, pr2 b(%20.3f) se(%12.3f) pr2(%9.3f)  varwidth(33) modelwidth(8) label coeflabels(_cons "Constant" OAminority "Black or Latinx Judge" OAminority_deep_cites "Black or Latinx Judge X Deep Cites") star(* 0.05 ** 0.01 *** 0.001) order(femaleOA femaleOA_deep_cites OAminority OAminority_deep_cites authorities_deep) replace drop(yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016) mtitles("Model 1" "Model 2" "Model 3") nonumbers

***** Table DA2.2: NBR Model of Citations, Asian-American Judges Included with Dummy 

********** Model 1: No Interaction
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority authorities_deep j1_asian splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & OAdist_ct~=1 , vce(cluster circuit)
estimates store model1_OA_asian_bin
********** Model 2: Interacting Non-white Judge and Gender with Deep Cites
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority femaleOA_deep_cites OAminority_deep_cites authorities_deep j1_asian splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & OAdist_ct~=1, vce(cluster circuit)
estimates store model2_OA_asian_bin
********** Model 3: Interacting Non-white Judge with Deep Cites 
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority OAminority_deep_cites authorities_deep j1_asian splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & OAdist_ct~=1, vce(cluster circuit)
estimates store model3_OA_asian_bin
esttab model1_OA_asian_bin  model2_OA_asian_bin model3_OA_asian_bin using DA2_2_APSR_asian_binary.rtf, pr2 b(%20.3f) se(%12.3f) pr2(%9.3f)  varwidth(33) modelwidth(8) label coeflabels(_cons "Constant" OAminority "Black or Latinx Judge" OAminority_deep_cites "Black or Latinx Judge X Deep Cites") star(* 0.05 ** 0.01 *** 0.001) order(femaleOA femaleOA_deep_cites OAminority OAminority_deep_cites authorities_deep j1_asian) replace drop(yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016) mtitles("Model 1" "Model 2" "Model 3") nonumbers


**** DA3. Alternative Specifications with Controls for Opinion Length. 

****** Correlations and Regressions establishing the correlations between Deeep Cites & Points of Law with exluded measures of opinion length and word count
corr wc opinlength authorities_deep keynotes if est_sample==1
reg  opinlength keynotes  authorities_deep  if est_sample==1
reg  wc keynotes authorities_deep  if est_sample==1

****** Table DA3.1: NBR Model of Citations, Control for Page Length of Majority Opinion
****** Model 1: No Interaction
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct opinlength yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)
estimates store model1_opinlength
****** Model 2: Interacting Non-white Judge and Gender with Deep Cites
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority femaleOA_deep_cites OAminority_deep_cites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct opinlength yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
estimates store model2_opinlength
****** Model 3: Interacting Non-white Judge with Deep Cites 
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority OAminority_deep_cites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct opinlength yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
estimates store model3_opinlength

esttab model1_opinlength model2_opinlength model3_opinlength using DA3_1_table_opinlength.rtf, pr2 b(%20.3f) se(%12.3f) pr2(%9.3f)  varwidth(27) modelwidth(8) label coeflabels(_cons "Constant") star(* 0.05 ** 0.01 *** 0.001) order(femaleOA femaleOA_deep_cites OAminority OAminority_deep_cites authorities_deep) replace drop(yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016) mtitles("Model 1" "Model 2" "Model 3") nonumbers

****** Table DA3. 2: NBR Model of Citations, Control for word count

********** Model 1: No Interaction
****** Model 1: No Interaction
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct wc yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)
estimates store model1_wc
****** Model 2: Interacting Non-white Judge and Gender with Deep Cites
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority femaleOA_deep_cites OAminority_deep_cites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct wc yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
estimates store model2_wc
****** Model 3: Interacting Non-white Judge with Deep Cites 
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority OAminority_deep_cites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct wc yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
estimates store model3_wc

esttab model1_wc model2_wc model3_wc using DA3_2_table_wc.rtf, pr2 b(%20.3f) se(%12.3f) pr2(%9.3f)  varwidth(27) modelwidth(8) label coeflabels(_cons "Constant") star(* 0.05 ** 0.01 *** 0.001) order(femaleOA femaleOA_deep_cites OAminority OAminority_deep_cites authorities_deep) replace drop(yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016) mtitles("Model 1" "Model 2" "Model 3") nonumbers

**** DA4. Additional Alternative Specifications for Federalist Society Associations

****** Table DA4.1: NBR Model of Citations, Control for Federalist Society Contributors
********** Model 1: No Interaction
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority authorities_deep OA_fed_soc_contributor splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 , vce(cluster circuit)
estimates store model1_fedsoc_contrib
********** Model 2: Interacting Non-white Judge and Gender with Deep Cites
 nbreg citation_5yr_diff_circ_2022 femaleOA OAminority femaleOA_deep_cites OAminority_deep_cites authorities_deep OA_fed_soc_contributor splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
estimates store model2_fedsoc_contrib
********** Model 3: Interacting Non-white Judge with Deep Cites 
 nbreg citation_5yr_diff_circ_2022 femaleOA OAminority OAminority_deep_cites authorities_deep OA_fed_soc_contributor splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)
estimates store model3_fedsoc_contrib

esttab model1_fedsoc_contrib  model2_fedsoc_contrib model3_fedsoc_contrib using DA4_fedsoc_contrib_APSR.rtf, pr2 b(%20.3f) se(%12.3f) pr2(%9.3f)  varwidth(27) modelwidth(8) label coeflabels(_cons "Constant") star(* 0.05 ** 0.01 *** 0.001) order(femaleOA femaleOA_deep_cites OAminority OAminority_deep_cites authorities_deep) replace drop(yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016) mtitles("Model 1" "Model 2" "Model 3") nonumbers



**** DA5. Alternative Specification Omitting Outliers 

****** Figure DA5.1: Distribution of Out-of-Circuit Subsequent Citations
qui nbreg citation_5yr_diff_circ_2022 i.femaleOA OAminority authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct i.year if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1 & OAdist_ct~=1, vce(cluster circuit)

histogram citation_5yr_diff_circ_2022 if e(sample), scheme(s2mono) graphregion(fcolor(gs16) lcolor(black))

graph save APSR_DA5_hist_DV.gph, replace
graph export APSR_DA5_hist_DV.tiff, as(tif) replace

****** Table DA5.1: NBR of Subsequent Cites, Cases with 10 or Fewer Subsequent Cites
********** Model 1: No Interaction
qui nbreg citation_5yr_diff_circ_2022 femaleOA OAminority authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1 & OAdist_ct~=1 & citation_5yr_diff_circ_2022<11, vce(cluster circuit)
estimates store model1_outlier

********** Model 2: Interacting Non-white Judge and Gender with Deep Cites
qui nbreg citation_5yr_diff_circ_2022  femaleOA OAminority femaleOA_deep_cites OAminority_deep_cites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1 & OAdist_ct~=10 & citation_5yr_diff_circ_2022<11, vce(cluster circuit)
estimates store model2_outlier

********** Model 3: Interacting Non-white Judge with Deep Cites 
qui nbreg citation_5yr_diff_circ_2022 femaleOA OAminority OAminority_deep_cites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 & citation_5yr_diff_circ_2022<11, vce(cluster circuit)
estimates store model3_interact

esttab model1_outlier  model2_outlier model3_interact using table_DA5_outliers_APSR.rtf, pr2 b(%20.3f) se(%12.3f) pr2(%9.3f)  varwidth(27) modelwidth(8) label coeflabels(_cons "Constant") star(* 0.05 ** 0.01 *** 0.001) order(femaleOA femaleOA_deep_cites OAminority OAminority_deep_cites authorities_deep) replace drop(yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016) mtitles("Model 1" "Model 2" "Model 3") nonumbers

****** DA6: Alternative models assessing trends over time
**** DA6.1: Tables of models used to generate the Figure OA5.1: Coefficient Plots of the Marginal Effects of Race/Gender on Subsequent Cites, by Year 
* gen yr_count_female=yr_count*femaleOA
* gen yr_count_minority=yr_count*OAminority
* gen yr_count_minority_women=yr_count*OA_minority_women
* gen yr_count_minority_man=yr_count*OA_minority_man
* gen yr_count_white_woman=yr_count*OA_white_woman
* lab var yr_count_female "Female Judge x Year Counter"
* lab var yr_count "Year Counter"
* lab var yr_count_minority "Non-white Judge x Year Counter"
* lab var yr_count_minority_man "Non-white Male Judge x Year Counter"
* lab var yr_count_minority_women "Non-white Female Judge x Year Counter"
* lab var yr_count_white_woman "White Female Judge x Year Counter"


nbreg citation_5yr_diff_circ_2022 i.femaleOA##c.yr_count i.OAminority##c.yr_count authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct  if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)

nbreg citation_5yr_diff_circ_2022 femaleOA yr_count_female OAminority yr_count_minority yr_count  authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct  if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)

estimates store yrcount_model1

nbreg citation_5yr_diff_circ_2022 i.OA_white_woman##c.yr_count i.OA_minority_women##c.yr_count i.OA_minority_man##c.yr_count authorities_deep  splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)

nbreg citation_5yr_diff_circ_2022 OA_minority_women  yr_count_minority_women OA_minority_man yr_count_minority_man OA_white_woman yr_count_white_woman yr_count  authorities_deep  splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1, vce(cluster circuit)

estimates store yrcount_model2_intersect
esttab yrcount_model1  yrcount_model2_intersect using table_DA6_1_yrcount_APSR.rtf, pr2 b(%20.3f) se(%12.3f) pr2(%9.3f)  varwidth(27) modelwidth(8) label coeflabels(_cons "Constant") star(* 0.05 ** 0.01 *** 0.001) order(femaleOA yr_count_female OAminority yr_count_minority OA_minority_women yr_count_minority_women OA_minority_man yr_count_minority_man OA_white_woman yr_count_white_woman yr_count authorities_deep) replace  mtitles("Model 1" "Model 2") nonumbers

************ Table DA6.2: NBR of Subsequent Cites: Splitting Data into Four-year Subsets
**** Analysis 2009-2012
****** Model 1: No Interaction
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 & year<2013, vce(cluster circuit)
estimates store OA10_m1_1st4

****** Model 2: Interacting Non-white Judge and Gender with Deep Cites
 nbreg citation_5yr_diff_circ_2022 femaleOA OAminority femaleOA_deep_cites OAminority_deep_cites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 & year<2013, vce(cluster circuit)
 estimates store OA10_m2_1st4
****** Model 3: Interacting Non-white Judge with Deep Cites 
 nbreg citation_5yr_diff_circ_2022 femaleOA OAminority OAminority_deep_cites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 & year<2013, vce(cluster circuit)
 estimates store OA10_m3_1st4

**** Analysis 2013-2016
****** Model 1: No Interaction
nbreg citation_5yr_diff_circ_2022 femaleOA OAminority authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 & year>2012, vce(cluster circuit)
 estimates store OA10_m1_2nd4

****** Model 2: Interacting Non-white Judge and Gender with Deep Cites
 nbreg citation_5yr_diff_circ_2022 femaleOA OAminority femaleOA_deep_cites OAminority_deep_cites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 & year>2012, vce(cluster circuit)
estimates store OA10_m2_2nd4

****** Model 3: Interacting Non-white Judge with Deep Cites 
 nbreg citation_5yr_diff_circ_2022 femaleOA OAminority OAminority_deep_cites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln  affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1 & year>2012, vce(cluster circuit)
estimates store OA10_m3_2nd4

esttab OA10_m1_1st4 OA10_m2_1st4 OA10_m3_1st4 OA10_m1_2nd4 OA10_m2_2nd4 OA10_m3_2nd4 using DA6_Table2_APSR.rtf, pr2 b(%20.3f) se(%12.3f) pr2(%9.3f)  varwidth(27) modelwidth(5) label coeflabels(_cons "Constant") star(* 0.05 ** 0.01 *** 0.001) order(femaleOA femaleOA_deep_cites OAminority OAminority_deep_cites authorities_deep) replace drop(yr_2010 yr_2011 yr_2012 yr_2014 yr_2015 yr_2016) mtitles("Model 1: 2009-12" "Model 2: 2009-12" "Model 3: 2009-12" "Model 1: 2013-16" "Model 2: 2013-16" "Model 3: 2013-16") nonumbers

************ Table DA6.3: NBR of Subsequent Cites, Splitting Data into Four-year Periods, Race-Gender
**** 2009-2012
nbreg citation_5yr_diff_circ_2022 OA_white_woman OA_minority_women OA_minority_man authorities_deep  splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1  & year<2013, vce(cluster circuit)
estimates store OA10_inter1_1st4

nbreg citation_5yr_diff_circ_2022  OA_minority_women OA_minority_man OA_white_woman OAminorwomen_deepcites OAminormen_deepcites OAwhitewomen_deepcites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1  & year<2013, vce(cluster circuit)
estimates store OA10_inter2_1st4


nbreg citation_5yr_diff_circ_2022  OA_minority_women OA_minority_man OA_white_woman OAminorwomen_deepcites OAminormen_deepcites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1  & year<2013, vce(cluster circuit)
estimates store OA10_inter3_1st4

*** 2013-2016
nbreg citation_5yr_diff_circ_2022 OA_white_woman OA_minority_women OA_minority_man authorities_deep  splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1  & year>2012, vce(cluster circuit)
estimates store OA10_inter1_2nd4
nbreg citation_5yr_diff_circ_2022  OA_minority_women OA_minority_man OA_white_woman OAminorwomen_deepcites OAminormen_deepcites OAwhitewomen_deepcites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1  & year>2012, vce(cluster circuit)
estimates store OA10_inter2_2nd4


nbreg citation_5yr_diff_circ_2022  OA_minority_women OA_minority_man OA_white_woman OAminorwomen_deepcites OAminormen_deepcites authorities_deep splitDemRep OA_reputation OAls_elite OAtenure_ln  AMICUS priorpubnum constit  dissent2 concur2 totalattys_ln termmeritsper_activejudge_ln affirm_and_reverse reverse multidoc_dum crim civ_lib_right keynotes readability_Nelson_Hinkle sixltr_pct yr_2014 yr_2015 yr_2016 if TREAT ~=9 & TREAT ~=10 & OPINSTAT==1 & j1_asian~=1  & OAdist_ct~=1  & year>2012, vce(cluster circuit)
estimates store OA10_inter3_2nd4


esttab OA10_inter1_1st4  OA10_inter2_1st4 OA10_inter3_1st4 OA10_inter1_2nd4 OA10_inter2_2nd4 OA10_inter3_2nd4 using DA6_Table3_APSR.rtf, pr2 b(%20.3f) se(%12.3f) pr2(%9.3f)  varwidth(27) modelwidth(5) label coeflabels(_cons "Constant") star(* 0.05 ** 0.01 *** 0.001) order(OA_minority_women OA_minority_man OA_white_woman OAminorwomen_deepcites OAminormen_deepcites OAwhitewomen_deepcites authorities_deep) replace drop(yr_2010 yr_2011 yr_2012 yr_2013 yr_2014 yr_2015 yr_2016) mtitles("Model 1: 2009-12" "Model 2: 2009-12" "Model 3: 2009-12" "Model 1: 2013-16" "Model 2: 2013-16" "Model 3: 2013-16") nonumbers
